Allogeneic transplantation, the only treatment for many genetic disorders, is limited by the lack of donors, toxicity of conditioning/immunosuppression, and graft vs. host disease (GVHD). Gene transfer in autologous hematopoietic stem cells (HSC) has shown promise, but long-term correction has been difficult to achieve due to inefficient HSC transduction and immune responses abrogating gene expression. Thus, employing gene transfer strategies to improve allogeneic transplant outcomes may prove to be a powerful alternative. Our overall goals are to develop safer protocols for allogeneic transplantation for genetic disorders using a novel positive/negative selection approach. Donor HSC will be endowed with a competitive advantage by transduction with tricistronic lentiviral vectors containing P140K-O6-methylguanine-methyltransferase (MGMTP140K), HSV-Thymidine kinase (TKHSV) and eGFP. Expression of these vectors enables in vivo chemo selection (positive selection) at the stem cell level by conferring resistance to benzyl guanine (BG), an inhibitor of endogenous MGMT but not MGMTP140K, and to chloroethylating agents such as BCNU. Expression of TKHSV with Ganciclovir (GCV) administration enables depletion (negative selection) of donor T cells mediating GVHD, and elimination of potential malignant clones arising from insertional mutations. Our preliminary studies establish proof-of-concept for in vivo selection of allogeneic donor cells in neonates, without myeloablation or immunosuppression, and for GCV mediated donor cell depletion without graft ablation. The focus of the current, revised R01 application is to apply this positive/negative selection approach in pre-clinical allogeneic murine models and in translational large animal studies. The hypotheses to be tested are: 1) non-ablative conditioning with busulfan or BG/BCNU, and post-transplant BG/BCNU treatment, will enable engraftment and in vivo expansion of allogeneic HSC, and suppress allo-reactive, untransduced cells of donor and host origin; 2) GCV administration will deplete transduced allo-reactive donor T cells causing GVHD and mitigate the need for post-transplant immunosuppression. To eliminate confounding variables of allo-immune responses, conditioning protocols will initially be developed in syngeneic murine models (revised AIM1A) and in autologous transplants in non-human primates (revised AIM 3). With the recent award of an NHLBI pilot project (January, 2011) (Center for Fetal Monkey Gene Transfer, UC Davis), non-human primate studies are now beginning, that will continue in the current AIMS. To enhance the translational impact of these studies, an EF1? directed-MGMTP140K/TKHSV vector (incorporated in a Phase 1 trial) will be tested (revised AIMs 1B-C). Critical parameters, required to apply MGMTP140K/TKHSV-based selection in allo-transplants, will then be examined in an MHC-mismatched murine model (AIM 2). Successful application of this positive/negative selection approach would be transformative for the field, combining advantages of allogeneic HSC carrying therapeutic genes in their normal genomic context, with the use of gene transfer methods to mitigate risks of cytoablation and GVHD.